I was surprised to open up shop this morning and find yields and MBS right about where we left them yesterday afternoon. I was also surprised yesterday afternoon when yields were able to hold under 1% by the end of the day. In other words, this bond rally has not only been impressively large, but also impressively resilient. Does that mean traders know where they want rates to go and are just taking their time (relatively) getting there?
No, I don't think so. I think yesterday was a factor of disillusionment with the Fed's rate cut, or perhaps a wake up call to certain investors due to the rate cut. The bond market has the complicated task of balancing the following sources of demand:
- temporary holding area for panicked cash seeking a sideline
- short positions seeking to cover to avoid taking any more of an unexpected beating than shorts already have
- legitimately pricing-in the longer-term economic impacts of coronavirus
- legitimately pricing-in a slower economy in the next presidential term, regardless of the winner
- legitimately pricing-in future friendly Fed policy
- legitimately pricing-in the excess fixed-income demand from foreign buyers whose own sovereigns are even deeper into negative territory (why did 10yr yields remain above 1% for so long in the first place?! -- only half joking)
- legitimately pricing in the temporary technical/algorithm-driven/headline-driven coronavirus momentum trade
- Asset reallocation trades from consumers shifting their 401k balances (forces money managers to buy bonds). Do you know many people talking about buying the dip in stocks right now?
- speculative traders who continue to buy when it seems that the herd mentality is that "we must be facing a big bounce soon"
- And on a more esoteric note, an ongoing massively long-term technical correction to rates that were way higher than they ever should have gotten at the end of 2018
For all these reasons and more, intraday yield movement is hard to pin down to one motivation. Even if we're only looking at the Fed as a motivation, it's complicated to sort out given the stock vs bond performance. This chart suggests stocks were more resilient due to the perception of friendly Fed policy, but also lost ground due to Fed disillusionment.
One decent explanation here is that yesterday's rate cut was indeed disillusioning, but that the market also generally expects more from the Fed, likely in the form of asset purchases some time in 2020 (or at the first major sign of economic slippage). I would assume the coronavirus economic impact gives the Fed great cover to justify such a move without any significant criticism.
But of course, all of the other market movement motivations are on the table as well. The only thing we can say for sure is that--regardless of motivation--levels are absolutely crazy. Essentially every lender is easily below their previous all-time rate lows from 2012. Levels are crazy enough and the move has been impressive enough that there's more than a small amount of concern for the good times coming to an end--or at least leveling off.
That combination (super low rates and fear of a bounce) has produced a refi boom like no other. Granted, outright counts of refi activity have been higher in the past, but never before have we seen such consistent acceleration in refi demand (even if that acceleration was made possible by refi levels being extremely low at the end of 2018). The following chart tracks the cumulative rate of change in refi apps. In other words, if apps grow 5% this week and 7% next week, the line moves to 12. If they decline 3% the following week, the line moves to 9%, and so on. The big totals in 2012 also saw big periodic pull-backs, which kept growth more even-keeled. This refi boom crushes that...
To be fair, I wouldn't necessarily read this chart as "higher = crazier." More like: the more vertical the move and the more ground it covers, the bigger the onslaught on the mortgage market's capacity to handle refi demand. In that sense, this is the first time since 2008/2009 that there has been any attack on that capacity worth mentioning (2001/2002 was big as well, but didn't make it on this chart).